The distinct effect of titanium dioxide nanoparticles in primary and immortalized cell lines

Abstract

The titanium dioxide nanoparticles (NPs) have been applied to biomedical, pharmaceutical, and food additive fields. However, the effect on health and the environment are conflicting; thus, it has been reviewing several times. In this context, establishing standard robust protocols for detecting cytotoxicity and genotoxicity of nanomaterials became essential for nanotechnology development. The cell type and the intrinsic characteristics of titanium dioxide NPs can influence nanotoxicity. In this work, the cyto- and genotoxicity effects of standard reference material titanium dioxide NPs in primary bovine fibroblasts and immortalized Chinese hamster ovary epithelial (CHO) cells were determined and compared for the first time. Titanium dioxide NPs exposure revealed no cytotoxicity for primary bovine fibroblasts, while only higher concentrations tested (10 μg/ml) induce genotoxic effects in this cell model. In contrast, the lower concentrations of the titanium dioxide NPs cause the cyto- and genotoxic effects in CHO cells. Therefore, our finding indicates that the CHO line was more sensitive toward the effects of titanium dioxide NPs than the primary bovine fibroblast, which should be valuable for their environmental risk assessment.

Graphical Abstract

Graphical Abstract.

Introduction

The titanium dioxide nanoparticles (TiO₂NP) are widely used in many products nowadays due to the unique properties, although there is no agreement on the effects in different cells. Among nanoparticles (NPs) of metal oxide, the TiO₂NP has been manufactured at high production volumes worldwide, which is expected to reach ~2.5 million tons by 2025 [1]. The TiO₂NP has been widely used in various industrial products such as cosmetics, sunscreens, toothpaste [2], dyes [3], and authorized food additive [4]. In the biomedical field, TiO₂NP can produce antimicrobial materials for the drug release vectors [5] and cancer treatment [6, 7], while, in the agriculture area, TiO₂NP has been reported to promote photosynthesis in some crops and disease control [8, 9].

The high applicability of TiO₂NP is due to the low dissolution rate, high surface area, anticorrosive, photocatalytic properties, low cost, abundance, and stability [10, 11]. However, evaluating potential risks for human health and the environment derived from exposure to nanomaterials is a matter of social/economic concern these days. The extensive use of TiO₂NP associated with improper disposal of domestic and industrial effluents may result in the release of this NP into the environment and subsequent impacts the ecosystem health. Some studies have been shown the toxic effect of TiO₂NP on organisms used as biomarkers of environmental risk [12–14]. It was demonstrated by Xia et al. [12] that TiO₂NPs caused oxidative damage, neurotoxicity and necrosis on the scallops. Bobori et al. [13] showed that oxidative, proteolytic, genotoxic, and apoptotic parameters increased due to TiO₂NPs exposure on land snail, prussian carp, and zebrafish organisms. Besides, Mylona et al. [14] reported that TiO₂NPs impair cell viability and growth of seagrass. In this context, the diversity and number of nanomaterials emerging in the market require novel approaches to assess their risks [9], understanding of the environmental, health, and safety aspects of TiO₂NP is essential to ensure the safe development of products containing these NPs.

According to how the nanomaterials were produced, the composition, charge, morphology, and surface area may influence the toxicity [15]. Most of the studies that assess nanotoxicity follow different standards (such as cell culture conditions and protocol), and the potential cell damage depends on various nanomaterials characteristics [15, 16]. In general, the nanomaterials used in the toxicological study are commercialized or synthesized in different laboratories. Thus, it is necessary to adopt standardized measurement methods and certified reference nanomaterials to assess their cytotoxic and genotoxic impact [17]. Besides, due to the unique characteristics of nanomaterials, several test guidelines and guidance documents for assessment of physicochemical properties, fate, exposure, and effects used for conventional chemicals need to be adapted when applied to nanomaterials [18].

Several in vitro toxicological studies using TiO₂NPs have been carried out in different cell types indicating distinct effects. Depending on the cell line and the protocol conditions, the cells may respond differently to the nanomaterials’ toxicity. The cell lines differed significantly in their response depending on nanomaterials type and toxicity endpoint measured [19]. Considering some in vitro toxicological studies using TiO₂NPs that evaluated the cell damage, the results have been showing conflict, which might be due to the variations in the cell type, route, and experimental conditions [20]. Some studies found that TiO₂NPs induce cyto- and genotoxic [21, 22]. In contrast, other reports showed that TiO₂NPs were not toxic to some cell lines [23, 24]. The cell features might explain these conflicting data, such as variation in cell physiology, proliferation state, membrane characteristics, and phagocyte characteristics among cell types [25].

Among nanotoxicity studies published, most of them use as experimental models immortalized cells or cancer cell lines [26]. Chinese hamster ovary epithelial (CHO) is in vitro model widely used in toxicological assays, suggests in Organization for Economic Co-operation and Development (OECD) genotoxicity guidelines, and is preferentially chosen due to tolerance to genetic manipulations, ease of adaptation, and rapid growth rates [27]. Primary bovine fibroblasts are a useful in vitro model to study the potentially toxic effects of nanomaterials exposure on animal production systems. The bovines are the most animals produced globally, so the concerns in health, maintenance, and production are increasing; moreover, knowing the properties of TiO₂NP makes it a great candidate to be used in bovine programs livestock. The TiO₂NP has been used as photocatalysts with antimicrobial activity in animal food additives to replace the feeds’ antibiotics [28]. According to some studies, the primary fibroblast indicates toxicity assessment of metallic NPs [26, 29]. To the best of our knowledge, no reports have investigated and compared the cyto- and genotoxicity of TiO₂NP in bovine fibroblast and CHO cell models.

Therefore, the present study aimed to investigate the in vitro cytotoxicity and genotoxicity of reference material TiO₂NP (NM01001a, European Union) using primary bovine fibroblasts and CHO cells. This study gathered the effect of TiO₂NP in a model of cells widely used in toxicity studies and another new cell line model of the farm animals. The results showed the difference between cyto- and genotoxicity effects of TiO₂NP in these cells and bringing a new perspective of applying this nanomaterial in animal production systems and environmental risk assessment.

Materials and Methods

Materials

The TiO₂NPs (NM01001a, European Union) used are a reference material sent to inter-laboratory studies of the NANoREG project, supplied by the Joint Research Centre. The supplier previously characterized these NPs. They had a primary particle size of 5–6 nm, particle size distribution 95% less than the 100 nm, 77% less than the 50 nm, and 11% less than the 10 nm, the surface area of 170/316 m [2]/g, and are anatase type [30]. For the cell culture, was used the Dulbecco’s modified Eagle’s medium (DMEM) (Sigma Aldrich, USA), Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F12) (Sigma Aldrich, USA) medium supplements fetal bovine serum (CAS n° 9014-81-7) (LGC Biotechnology, Brazil) (10%) and penicillin–streptomycin antibiotics (LGC Biotechnology), phosphate-buffered saline without calcium and magnesium (PBS, CAS n°1314-87-0) (LGC Biotechnology) and trypsin–EDTA (CAS n° 9002-07-7). For biological analyzes, were used the thiazolyl blue tetrazolium bromide (MTT, CAS n° 298–93-1) (Sigma Aldrich, USA), superoxide dismutase (SOD) assay kit (Sigma Aldrich, USA; 19160), cyclophosphamide (CAS n°6055-19-2) (Sigma Aldrich, USA), colchicine (CAS n°64–86-8) (Sigma Aldrich, USA), methyl methanesulfonate (MMS, CAS n° 66–27-3) (Sigma Aldrich, USA), Cytochalasin β (Cyto β, CAS n°14930–96-2) (Sigma Aldrich, USA), metabolic activation system (S9 mix) (Sigma Aldrich, USA; S2067), and Giemsa (Sigma Aldrich, USA) (CAS n°51811–82-6).

Dispersion and characterization of NPs

The characterization assays are essential in nanotoxicity studies since it predicts the interaction mechanism in biological systems, so we performed the chemical composition, the morphology, hydrodynamic size, and Zeta potential. For all analyses, the TiO₂NPs (1000 μg/ml) were sonicated for 16 min in the Branson Sonifier S-450 (Branson Sonifier S-450, USA) at 400Watts, based on NANoREG—the nanogenotox dispersion protocol [31]. The chemical composition of TiO₂NP was evaluated using Bomem Fourier-transform infrared (FTIR) spectrometer MB-102 (Bomem, Canada). The analysis was then performed in the 4000 to the 200 cm⁻¹ region using KBr pellets previously dried at 500°C. Raman measurements were performed on a Bruker RFS 100 (Bruker, USA) spectrometer excited with an Nd + 3/YAG laser at 1064 nm, equipped with liquid nitrogen cooled InGaAs detector. The spectral were acquired at 4-cm⁻¹ resolutions, and an average of 512 scans was collected with a laser power of 80 mW.

The morphology of TiO₂NP was evaluated using an atomic force (AFM) (Nanosurf EasyScan 2, Nanosurf Instruments, Switzerland). For this analysis, the NPs were added to an isopropyl alcohol/ultra-pure water dispersion and exposed to ultrasonic waves. Before analysis, a drop of the dispersion was placed on mica’s surface and then dehydrated under the nitrogen atmosphere. The analyses were performed at cantilever free oscillation amplitudes (50 mV), and the size of the scanning area of the microscope probe was 5 × 5μm. The AFM resolution was 512 × 512 pixels, and the tip length and radius were 225 μm and <10 nm, respectively.

Furthermore, we evaluated the hydrodynamic size, formation of aggregates, and Zeta potential of TiO₂NPs using the dynamic light scattering (DLS) technique by the Malvern 3000 Zetasizer NanoZS equipment (Malvern, UK). The analyses were done at concentrations of 0.1, 1, 5, and 10 μg/ml of TiO₂NP in three different media: deionized water, DMEM-F12, and DMEM-F12 with fetal bovine serum previous sonicated. All media tested had a pH of 7.0.

Cell line culture and exposure to NPs

To determine if the TiO₂NP may cause cyto- and genotoxicity that was used in vitro model, it allows the study of the primary effect in cells. Before in vitro analyses, the TiO₂NPs (1000 μg/ml) were sonicated for 16 min in the Branson Sonifier S-450 (Branson, USA) at 400 Watts, based on NANoREG—the nanogenotox dispersion protocol [31], and subsequently, the treatments were prepared at concentrations of 0.1, 1, 5, and 10 μg/ml. These concentrations were chosen because some studies using low concentrations of nanomaterials showed damage in cells [32, 33]. Thus, they are potentially more realistic (due to low concentrations). This range of TiO₂NP concentration is less studied.

The CHO and primary bovine fibroblasts generated from a piece of tissue (1 cm [2]) extracted from the Gir cow ear were grown into the T-25 flasks in DMEM/F12 and DMEM culture medium, respectively. Both culture mediums were supplemented with 10% (v/v) of fetal bovine serum and 1% (v/v) antibiotic (100-IU/ml penicillin, 100-μg streptomycin/ml) and incubated at 37°C in a humidified atmosphere containing 5% CO₂ (Model 4130, Thermo Fisher Scientific, Waltham, MA, USA). All cytotoxic assays were done after exposure to various concentrations (0—negative control; 0.1, 1, 5, and 10 μg/ml) of TiO₂NP dispersed in the culture medium and incubated for 24 h. Three independent experiments were done, each carried out in triplicate.

The exposure to TiO₂NP may result in cell morphology changes, so comparative evaluations between control and treatments contribute to toxicity analyses with a direct indicator in the evaluation of cytotoxicity. Thus, this analysis was performed under an inverted light microscope (ICM 405, Zeiss, Germany) with a camera (Axiocam ERC 5 s, Zeiss, Germany) attached to a computer for the images to be captured in the Zen 2.3 lite software.

Methyl tetrazolium assay

The cell viability determination through the mitochondrial activity was carried out with methyl tetrazolium (MTT) assay based on ISO 10993-5 (ISO2009). Briefly, 6 × 10 [3] cells/well for the bovine fibroblasts and 12 × 10 [3] cells/well for the CHO were plated in 96-well plates. After 24 h of NP exposure, the medium was removed, and the cells incubated with 10% (v/v) MTT in culture medium for 4 h. This medium was then removed for the addition of Isopropanol-Acid (0.04 M HCl) to dilute the formazan formed, so the samples were read on the spectrophotometer (Varioskan Flash Multimode Reader, Thermo Fisher Scientific, USA) at 570 nm. To avoid the potential nanomaterials interference with the MTT assay, the absorbance values ​​were corrected per well containing only TiO₂NPs.

SOD assay

Oxidative stress is a mechanism widely reported in nanotoxicity [34], and SODs are antioxidant enzymes that protect against these harmful biological events [35]. According to the manufacturer’s instructions, the SOD activity was determined via a spectrophotometric method using an SOD assay kit (Sigma Aldrich, USA; 19160). Briefly, 2 × 10 [4] cells/well were seeded into 24-well plates containing culture medium and maintained in a humidified, 5% CO₂ incubator at 37°C. After 24 h of TiO₂NP exposure, the cells were trypsinized and lysed by freezing and thawing. Then, 60 μl of PBS was added, samples were centrifuged at 2000 g for 5 min, 4°C, and 20 μl of the supernatant was used to proceed to assay according to the SOD analysis protocol. The samples were read on the spectrophotometer (Varioskan) at 450 nm.

Micronucleus assay

Since the nanomaterials may cause mutagenic activity and induce micronuclei formation, micronucleus assay was performed using the OECD TG-487 [36]. The cells (1 × 10 [5] cells/well) were in vitro cultured into 6-well plates in a humidified, 5% CO₂ incubator at 37°C. Throughout the experiment, the cells were exposed to various treatments. The scheme of the exposition period in micronucleus assay is summarized in Fig. S1, see online supplementary material for a colour version of this figure. In the short treatment of 3 h, the groups of the presence of an S9 mix (1% v/v) are cyclophosphamide (5 μg/ml, positive control), 0.1, 1, 5, and 10 μg/ml of TiO₂NP. As well as the groups without the S9 mix, which are: MMS (13 mg/ml, positive control), Colchicine (0.5 mg/ml, positive control), 0.1, 1, 5, and 10 μg/ml of TiO₂NP, and the negative control (0 μg/ml of TiO₂NP). After this exposure, the cells were maintained for 24 h in a culture medium with Cyto β (3 μg/ml). The Cyto β (3 μg/ml) was added to all treatments in the long treatment and contacted the cells for 24 h.

After the exposure period (3 h—short or 24 h—long), the cells were harvested and centrifuged at 120 g for 5 min. Hypotonic treatment consisted of careful resuspension of the cells in 300-μl hypotonic saline (0.075 M KCl). Immediately after adding the hypotonic solution, the cells were centrifuged for 5 min at 120 g. The fixation was then done to resuspend the pellet in 100-μl Carnoy I fixative (ethanol: acetic acid, 3:1 v/v), and then, the smear was performed. The slides were stained with Giemsa for 10 min. The numbers of micronucleus per 2000 binucleated cells were recorded.

Statistical analysis

Data from the MTT and SOD assays were submitted for analysis by the ANOVA. Then, the means were compared by the Tukey test. Micronucleus assay data were analyzed, comparing the groups treated with the control by the Chi-square test. Values of P < 0.05 were considered significant, and the numerical data were presented as the mean and SEM. Moreover, all data were analyzed in GraphPad software 5 (GraphPad Software, USA).

Results and Discussion

NPs characterization

Before the in vitro cell toxicity assays, we analyzed the chemical–physical characteristics of the TiO₂NPs. TiO₂NP was characterized using AFM, Raman, DLS, and infrared spectroscopy. These properties are essential for a better understanding of NP toxicity in vitro mammalian cell models. According to the Infrared Spectrum, three bands were pointed out (Fig. 1A). The large bands observed in the infrared spectrum at 542 and 686 cm⁻¹ are associated with the vibration of Ti-O bonds, which indicates that it is a titanium dioxide. The band at 3427 cm⁻¹ suggests hydroxylated surfaces caused by the hydrogen interactions between OH groups of titanium dioxide surface with +2 oxidation and molecular water with vibrations at 1630 cm⁻¹ [37, 38]. The bands observed in the Raman spectrum obtained in 639, 517, and 395 cm⁻¹ might be attributed to B_1g, A_1g/B_1g e E_g (Fig. 1B).

Figure 1 .

Characterization of titanium dioxide nanoparticles (NPs); (A) FTIR spectrum and (B) Raman spectrum of TiO₂NP; (C) three-dimensional Atomic Force Microscopy (AFM) images of titanium dioxide nanoparticles (TiO₂NP); the horizontal axis represents the x- and y-axes (μm), and the vertical axis represents the z-axis (μm).

The images obtained by the AFM show that TiO₂NPs are under various geometries and sizes, being outstanding in the cylindrical, cubic, and pyramidal forms (Fig. 1C). Beddoes, Case, and Briscoe [39] reported that the NP exhibits different translocation capacity inside the cells depending on the shape. In this way, the variety of shapes observed in the TiO₂NP may impact different ways in cell and toxic effects.

DLS measurements (Table 1) showed that the hydrodynamic sizes of TiO₂NPs varied between 500 and 1300 nm in the dispersions of deionized water and cell culture medium, respectively. However, the hydrodynamic size of TiO₂NPs, in cell culture medium with fetal bovine serum, decreased compared with the medium without fetal bovine serum, suggesting which the protein adsorption promotes better dispersion of TiO₂NPs (20.8 ± 0.1 nm for the lower and 444.8 ± 16 nm for the higher concentration tested).

Table 1.

data from DLS analyses in a different medium

Aqueous medium	TiO2NP (μg/ml)	Hydrodynamic Size (nm)	Polydispersity indices	Zeta potential (mV)
Deionized water	0.1	668.50 ± 32.10	0.526 ± 0.05	−3.50 ± 0.40
	1	576.50 ± 28.90	0.395 ± 0.02	−11.80 ± 0.10
	5	799.80 ± 408.00	0.854 ± 0.48	−0.10 ± 0.50
	10	1,246.70 ± 5.80	0.277 ± 0.02	−1.40 ± 0.04
DMEM/F12	0.1	1,330.30 ± 167.10	0.753 ± 0.11	−22.30 ± 0.70
	1	1,184.00 ± 44.50	0.530 ± 0.01	−20.60 ± 1.70
	5	1,145.00 ± 65.10	0.352 ± 0.08	−22.20 ± 1.40
	10	1,226.70 ± 5.80	0.266 ± 0.04	−21.40 ± 1.20
DMEM/F12 + fetal bovine serum	0.1	20.80 ± 0.10	0.534 ± 0.01	−15.80 ± 2.10
	1	57.60 ± 7.50	0.738 ± 0.14	−16.40 ± 1.50
	5	309.80 ± 33.40	0.751 ± 0.07	−14.20 ± 1.80
	10	444.80 ± 16.00	0.466 ± 0.06	−13.30 ± 1.40

Data are given as mean ± SEM.

After, we measured the Zeta Potential of the TiO₂NPs (Table 1). In deionized water, the NPs dispersions were unstable, with values close to zero. However, when the TiO₂NP was dispersed into the culture medium, there was an increase in Zeta Potential values, especially in DMEM/F12, without fetal bovine serum when the values were on average −21.62 mV. In some treatments, polydispersion index values were higher than the 0.7, as in 5 μg/ml (deionized water), 0.1 μg/ml (culture medium), and 1 and 5 μg/ml (culture medium with fetal bovine serum) of TiO₂NP, indicating heterogeneity in the sample (Table 1). Besides, in 10 μg/ml of TiO₂NP (in deionized water and culture medium), the values were from almost monodisperse populations. In the case of dispersions of 0.1 μg/ml (deionized water), 5 μg/ml (culture medium), 0.1 and 10 μg/ml (culture medium with fetal bovine serum) of TiO₂NP had a population with average polydispersity. The physicochemical characteristics of TiO₂NP, such as hydrodynamic diameter, polydispersity index, and Zeta Potential in different aqueous media and concentrations, are summarized in Table 1. The method chosen to test the TiO₂NP in different liquid dispersants reinforces the different behavior of these NPs.

The DLS analysis demonstrates aggregation that confirms the influence of the composition of the aqueous media in the NPs. This effect may have happened because of van der Waals interactions; permanent or induced dipoles within the NP can result in net attractive forces between NPs and subsequent aggregation [40]. To minimize NPs aggregation’s effects, dispersion of TiO₂NPs was always freshly prepared and sonicated before each experiment. In addition to sonication, it was showed that the fetal bovine serum supplementation in culture medium contributes to maintaining dispersed NPs in colloidal dispersion. Upon contacting the NPs into the culture medium, their surfaces will be covered by the serum proteins, forming the so-called “protein corona” [41, 42]. These proteins may hinder the interaction NP-NP by the Van der Wall forces that facilitate the dispersion of TiO₂NP. Also, the negative charge of TiO₂NP can interact with Ca²⁺ e Mg²⁺ through electrostatic forces [43]. Here, these processes can impact the formation of type and size of external layer surface TiO₂NPs. We observed smaller size values in the lowest concentrations of TiO₂NP, probably due to a more dispersed NP dispersion. The small size also provides a more exposed surface area of NP to the serum protein interaction [44]. Therefore, these factors can explain the DLS size results. The potential behaviors of TiO₂NP in the mediums are summarized in Fig. 2.

Figure 2 .

Representative scheme of titanium dioxide nanoparticles (TiO₂NP) aggregates in deionized water (A), culture medium (B), and culture medium supplemented with fetal bovine serum (C). Polar hydroxyl (-OH) groups can cause the surface to physically attract and absorb one or more additional layers of polar water molecules, which gives the oxide a hydrophilic character (A). The TiO₂NPs can interact with Ca²⁺ and Mg²⁺ ions in the culture medium by electrostatic forces (B). Protein albumin present in serum possibly interacts with the Ca²⁺ ion associated with TiO₂NPs by electrostatic forces, making it difficult for Van der Waals to interact with TiO₂NPs make it disperse (C). The illustration was done in Biorender.

In the present study, the Zeta Potential results show values varied among the different media tested. According to previous studies, this variation can be due to ions in the different aqueous media [45, 46]. In all aqueous solutions, the Zeta Potential was negative, and the highest values have been observed in the culture medium (with and without serum). Many metal oxides hydrolyze in the presence of water can accumulate hydroxide layers on the surface. The hydroxyl groups (-OH) may cause the surface to physically attract and absorb a single or multiple additional layers of polar water molecules, which confers a hydrophilic character to the oxide [47]. Through the infrared spectroscopy, the analyses confirmed which the TiO₂NP presents a hydrophilic character due to the band 3427 cm⁻¹ indicating the interaction of hydrogen of the OH groups with the surface of the titanium dioxide, forming several layers of water molecules. Besides, the ions present in the culture media may have contributed to the negative charge of TiO₂NP.

In vitro toxicity study

Optical microscopy observation was first used to evaluate the cytocompatibility of TiO₂NPs to different cell types. We evaluated fibroblast and CHO cells’ morphology after 24-h exposure to 0.1, 1, 5, and 10 μg/ml of TiO₂NPs. Fig. 3 shows that the TiO₂NP did not alter both cells’ morphology compared with the non-exposed control. In the high concentration, we noted TiO₂NPs aggregates (shown in the arrows in Fig. 3). These observations are in good agreement with the hydrodynamic size and the heterogeneity shown in DLS analysis. The correlation between the DLS and microscope observation is interesting because some studies reported that TiO₂NP aggregates could cause cytotoxic [48–50].

Figure 3 .

Morphological analysis of CHO (A-E) and bovine fibroblast (F-J) cells treated with titanium dioxide nanoparticles (TiO₂NP) after 24 h. Control group (A and F), .1 μg/mL (B and G), 1 μg/mL (C and H), 5 μg/mL (D and I) and 10 μg/mL TiO₂NP (E and J). The arrows show the aggregations. Magnification x200. Scale bar = 20 μm.

The cell viability of TiO₂NP was assessed by MTT assay (Fig. 4). We observed no changes in primary bovine fibroblasts cells than in control cells (Fig. 4A; P > 0.05). However, in CHO cells, only in the 0.1-μg/ml TiO₂NP after 24-h exposure, higher absorbance was observed (Fig. 4B; P < 0.05), indicating mitochondrial dysfunction activity. This result is in line with some studies that showed the TiO₂NPs exposure induced an effect on mitochondrial function [51, 52]. Simultaneously, van Vliet et al. [53] observed that the mitochondria could increase ATP production under cell stress to promote cell survival. There was likely due to a hormesis effect. In this condition, the cells accelerate the proliferation after low dose NPs exposition [54, 55]. Furthermore, the NPs may be interacting with cell receptors related to growth factors, cell cycle regulation (such as kinases), and the functioning of mitochondria [37, 56]. The method used to analyze the mitochondrial activity is an indirect measurement to predict cell proliferation. During the cell cycle, the mitochondria are essential in regulating and balancing fission and fusion, providing energy to duplicate the cell [57, 58]. Therefore, the increase noted in 0.1 μg/ml of TiO₂NP in CHO cell may indicate a rise in cell proliferation. Further data collection is required to determine precisely how a low concentration of TiO₂NP in CHO cells affects mitochondrial activity and whether it changes the cell cycle.

Figure 4 .

The mitochondrial activity determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay in bovine fibroblast (A) and CHO (B) cells cultured under different concentrations (.1, 1, 5 and 10 μg/mL) of titanium dioxide nanoparticles (TiO₂NPs) after 24 h. Data were analyzed in the Tukey test (P < .05). Different letters mean statistical differences.

As mitochondrial is responsible for the generation of free radicals [35], the formation of free radical-mediated by TiO₂NPs [59] was investigated the stress oxidative by a common biomarker of antioxidants defenses by SOD assay. These tests revealed that after 24 h of exposure, TiO₂NPs did not induce SOD activity changes (P > 0.05) for the fibroblast (Fig. 5A) and CHO (Fig. 5B) cells. Despite the observed change in CHO cell viability after the exposure to 0.1 μg/ml of TiO₂NP, we did not find changes in SOD activity after the same exposure condition.

Figure 5 .

Oxidative stress by superoxide dismutase (SOD) assay of bovine fibroblast (A) and CHO (B) cells cultured under different concentrations (.1, 1, 5, and 10 μg/mL) of titanium dioxide nanoparticles (TiO₂NPs) after 24 h. Data were analyzed in the Tukey test (P < .05).

Nanomaterials might produce reactive oxygen species (ROS) due to their large surface area [35, 60]. The possible mechanism of the toxicity of TiO₂NP is ROS’s action produced by these NPs after the induction of electron pairs under the light [61]. However, in the present study, no changes were observed in the activity of the SOD enzyme for both fibroblasts and CHO cells, despite noted the changes in mitochondrial activity in CHO cell.

Genotoxicity evaluation

The in vitro micronucleus test can detect chromosome fragments or lagging intact chromosomes in the cytoplasm of cells that have undergone division. Therefore, the evaluation of chromosome damages is an important tool of genetic toxicology because the chromosomal mutation is a determinant event in carcinogenesis. To assess the micronucleus frequency in the cells exposed to the TiO₂NPs, we used the high recommend OECD TG-487 Guideline [36]. The micronucleus assay results were determined after the in vitro exposition of NPs at 0.1, 1, 5, and 10 μg/ml, times (short or long), and with and without S9 (metabolic activation system) to primary bovine fibroblasts and CHO cell models.

Representative images of micronucleus in the fibroblast and CHO cells exposed to TiO₂NP are demonstrated in Fig. 6. For the CHO cell line, the TiO₂NP induced the micronucleus formations in low concentration tested (Table 2). The high values of micronucleus were found in short treatment when exposure 0.1 μg/ml of TiO₂NP (with and without S9 fraction) when compared with control (P < 0.05). Moreover, in the long treatment 1 μg/ml of TiO₂NP also increased the micronucleus frequency (Table 2, P < 0.05). It is interesting to note that the increase in time of exposure influences micronucleus formation. On the other hand, in primary bovine fibroblasts, no change in micronucleus formations was observed, even if in the same conditions of TiO₂NP exposure. However, a decrease in micronucleus frequency was observed only after exposition to 10 μg/ml of TiO₂NP in short treatment (Table 3, P < 0.05). Although this low micronucleus frequency, in this case, we believe that this result is due to the high number of aggregates in the slides that difficult the count. The mix S9 appears to be harmful to primary bovine fibroblasts since it caused cells to die when exposed.

Figure 6 .

Representative light microphotography illustrating the micronucleus (MN) assay. (A) Bovine fibroblast exposed to 5 μg/mL titanium dioxide nanoparticles (TiO₂NP), the arrow shows the micronucleus presence. (B) CHO cell exposed to 10 μg/mL TiO₂NP, the arrow shows the NP. (C) Bovine fibroblast exposed to methyl methanesulfonate (positive control), the arrow shows the nuclear bud. (D) Bovine fibroblast exposed to colchicine (positive control), the arrows show bridge and micronucleus. Giemsa staining. Magnification x400. Scale bar = 50 μm.

Table 2.

micronucleus frequencies in CHO cells after short (3 h) and long (24 h) exposure time to TiO₂NP

Groups	Micronucleus Frequencies (%)		CBPI (Cytokinesis-Block Proliferation Index)
	Short	Long	Short	Long
Negative control	0.55	0.55	1.86	1.87
0.1-μg/ml TiO2NP	1.30*	1.05*	1.80	1.84
1-μg/ml TiO2NP	0.85	1.05*	2.01	1.82
5-μg/ml TiO2NP	0.30	0.35	1.90	1.92
10-μg/ml TiO2NP	0.35	0.30	1.66	1.93
MMS (positive control)	1.40	1.10	1.78	1.87
Colchicine (positive control)	1.65	1.00	1.85	1.86
Cyclophosphamide (positive control)	2.15	1.10	1.74	1.85
0.1-μg/ml TiO2NP + S9	2.05*	−	1.96	−
1-μg/ml TiO2NP + S9	0.70	−	1.77	−
5-μg/ml TiO2NP + S9	0.65	−	1.77	−
10-μg/ml TiO2NP + S9	0.45	−	1.85	−

Data of frequencies were calculated in some micronucleus/2000 binucleated cells observed, analyzed comparing the groups treated with the control by chi-square test with a significance level of 5%. ^*P < 0.05 means a significant difference. - Means without data since cell death with S9 exposure.

Table 3.

micronucleus frequencies in bovine fibroblast cells after short (3 h) and long (24 h) exposure time to TiO₂NP

Groups	Micronucleus Frequencies (%)		CBPI
	Short	Long	Short	Long
Negative control	0.55	0.60	1.65	1.59
0.1-μg/ml TiO2NP	0.70	0.55	1.59	1.73
1-μg/ml TiO2NP	0.60	0.60	1.61	1.58
5-μg/ml TiO2NP	0.35	0.50	1.69	1.67
10-μg/ml TiO2NP	0.15*	0.55	1.58	1.77
MMS (positive control)	1.30	1.15	1.57	1.60
Colchicine (positive control)	1.05	1.20	1.62	1.67
Cyclophosphamide (positive control)	−	0.75	−	1.73

Data of frequencies were calculated in some micronucleus/2000 binucleated cells observed, analyzed comparing the groups treated with the control by chi-square test with a significance level of 5%. ^*P < 0.05 means a significant difference. - Means without data since cell death with positive control exposure.

The present findings suggest that the low concentration and smaller sizes of TiO₂NP could cause micronucleus formations. The micronucleus formation in CHO in the lower concentrations tested is in concurs well with the findings of Di Virgilio et al. [62], which show micronucleus formation in CHO after exposure of 0.5- and 1-μg/ml TiO₂NP.

As proposed in the literature, particle size indicates an essential role in genotoxicity [63, 64]. The small size NPs present higher surface reactivity that may induce interaction with endocytosis receptors [65]. Another study has indicated that the small size NPs (50 nm) are internalized more efficiently than the large NP [66]. According to the DLS analyses, the TiO₂NP exhibited an average size of 20.8 nm (0.1 μg/ml) and 57.6 nm (1 μg/ml) in the culture medium with serum. Under these conditions, the cells may have easily internalized the TiO₂NP. After cell uptake, the NP could be interacting by physical action with DNA, destabilizing their double helix structure [67, 68]. The TiO₂NPs are known for their high affinity with phosphate groups [69], possibly engendering a high reactivity and trapping against DNA itself. Other mechanisms of genotoxicity may be attributed to the physical interference of the TiO₂NPs in the DNA repair system or alteration of the mitotic spindle apparatus and chromosome segregation [17]. Therefore, in the present study, it is possible to hypothesize that the genotoxicity observed in the lowest TiO₂NP concentrations can occur because of NP’s high reactivity into well-dispersed NPs. Presumably, as mitochondrial dysfunction can affect the cell cycle, the TiO₂NP might are acting indirectly in this pathway by accelerating the cell cycle and disturbing the mitotic checkpoint, inducing the micronucleus formation. There is evidence to suggest that TiO₂NP deregulate the function of the polo-like kinase 1 protein from the mitotic checkpoint that controls various processes during mitosis, including contractile ring formation and cytokinesis [37].

Remarkably, in the present study, the immortalized cell shows to be sensitive than the primary cell line. The immortalized cells have modifications in cell cycle checkpoint pathways or in telomerase to facilitate their long-term cultivation in vitro; therefore, it might be the reason for genomic instability. According to observations on long-term culture, CHO cells’ genomic instability has related to a compromised DNA repair and transcription machinery, likewise a high occurrence of haploidy [27]. Hence, the different effects to NPs compared with primary cell lines may be related to intrinsic characteristics of immortalized cells.

These findings add to a growing body of literature on our understanding of the effect of the reference TiO₂NP occasioning different toxic responses in primary bovine fibroblasts and CHO cells. Although CHO cell has been suggested in the OECD guideline (TG-487), it is scary for the genotoxic effect of TiO₂NP in these cells. Our findings provided further evidence that CHO cell is a sensitive toxicological model since a low concentration of TiO₂NP causes dysfunction in mitochondrial activity and micronucleus formation. The bovine fibroblast cells’ results so far have been encouraging the application of TiO₂NP in agribusiness. The unique properties of TiO₂NP (such as efficient photocatalysis) and no toxic effect make it a great NP to be applied in cattle production and agribusiness.

In this research, we sought to investigate the effect of TiO₂NP in two line cells, and as expected, the TiO₂NP affects both in different ways. Our study provides further evidence that CHO cell shows more sensitivity to toxicity than primary bovine fibroblast after TiO₂NP exposure. Moreover, the lower concentration of TiO₂NP exhibits small hydrodynamic size; consequently, it may cause cyto- and genotoxicity effects.

Conclusion

As cell types may react differently in the same environment, we sought to investigate the effect of TiO₂NP in primary bovine fibroblast and CHO cells. The utilization of standardized toxicity protocols and reference nanomaterials is crucial for reliable risk assessments in the nanotoxicology field. Our finding confirms that the toxicological response to TiO₂NP also depends if the cell is primary or immortalized cell lines. The TiO₂NP did not affect the primary bovine fibroblasts cell line. However, the results reveal the lower concentrations higher susceptibility to cause cyto- and genotoxicity in immortalized CHO cells. In our view, these results are valuable for applications of the TiO₂NP in bovine programs livestock and environmental impacts assessment.

Conflict of interest statement

The authors report no conflicts of interest. The authors are solely responsible for the content and writing of this article.